Uwe Reinhardt on being a good economist

When I entered the field of health economics in the early 1980s, the prevailing wisdom within the field was different than what it is today:

• First, the idea that physicians could induce demand for their services seemed revolutionary. One notable exception was Uwe Reinhardt, who wrote in 1985 that there was really only one group of people in the whole world that found the idea revolutionary: American economists. ¹

• Second, universal coverage was not on economists’ policy table, with one prominent researcher pointing out that Dr Reinhardt was one of only a very few health economists who were vocal supporters of the idea. ² Currently, support for universal coverage appears to be the norm among American health economists, in that almost 90% support the Affordable Care Act and over 80% believe in the individual mandate. ³

• Third, it was generally held that there was so much moral hazard that national health insurance with low patient cost sharing would be a colossal waste of money. The principal investigators of the RAND Health Insurance Experiment, which cost about three‐quarters of a billion dollars in today's dollars, once claimed that, depending on the validity of certain (admittedly, somewhat heroic) assumptions, the entire cost of the experiment would have been recouped in a single week because employers and insurers increased cost sharing requirements considerably in the wake of the study's publication. ⁴ How could this be? They contend that it was because there was so much less moral hazard in the wake of the higher cost sharing requirements in the workplace. I think it's significant that this estimate ignores the fact that other parts of the study showed that cost sharing reduced the use of useful medical care as much as unnecessary treatment. ⁵

• And fourth, most economists held to the notion that providing free care was “inefficient”—with the natural corollary that the more people were charged for care, the more efficient our health care system would be because this would reduce moral hazard. I have two intellectual heroes: Uwe Reinhardt and Robert Evans from Canada, whom I consider the primary mentors in my career. Evans said it quite plainly—that if you believe that efficiency and higher cost sharing are positively correlated, then “The welfare burden is minimized when there is no insurance at all.” ^{6 (p. 49)}

  Following that line of inquiry, are we entitled to assume that the Canadian, government‐run health‐insurance system, which does offer patients first‐dollar health‐insurance coverage all around, is ipso facto less efficient than the American system which visits substantial cost‐sharing on patients and leaves millions of them uninsured to confront the full cost of the health care they consume? ^{7 (p. 303)}

  Reinhardt made this further tangible when he wrote,

I think it's safe to say that most health economists at the time I entered the field would have answered “yes” to that question—the Canadian system is by definition less efficient; and there was no need to look at the data because theory said so.

As a result, there was not a home in societies of economists for those questioning these orthodoxies when I entered the field around 1982. It was therefore very fortunate for me—and so many others—that there were visionaries like Stuart Altman, Clif Gaus, and Uwe Reinhardt who helped form and sustain the forerunner of AcademyHealth, providing us with not only an outlet for our research, but the camaraderie that one seeks in a professional society.

I believe that Dr Reinhardt was instrumental in changing the field's collective mind about these issues; they were hardly mainstream ideas at the time. He was a revolutionary, although sometimes it was hard to observe because he had a way of making people of every political persuasion feel so at ease.

I have titled my talk, “Uwe Reinhardt on Being a Good Economist.” It is based almost entirely on his writings between 1978 and 2001—that is, 20 to 40 years ago—the period where he was thinking and writing prolifically about welfare economics in general, and efficiency and equity in particular. The term “welfare economics” is not widely understood outside of the field; it examines what is the best way to organize an economic society. This involves both the use of and distribution of society's resources in the best manner.

Reinhardt wrote so much about welfare economics, efficiency, and equity during this period that it is hard to even know where to begin. What seemed to have bothered him the most was that economists who were very well trained, and who therefore should have known better, were guilty of a cardinal sin—although he was too generous to opine whether it was a sin of omission or a sin of commission.

In a nutshell, the cardinal sin is this: the study of microeconomics is all about efficiency, but economists use the term in a way that leads (or even dupes) an unsuspecting (but otherwise well‐informed) public into believing that unfettered price competition in general—and patient cost sharing in particular—is in society's best interest.

Reinhardt wrote a lot about efficiency during this period. He was particularly irritated by a proposal by Milton Friedman, the consummate free marketeer and Nobel Laureate. Friedman attacked the Medicare program, a full 25 years after its founding, as being so very inefficient that fixing it would entail replacing it and Medicaid with a program where every American would be required to have health insurance with a deductible of 30% of family income, which is about $20 000 in today's dollars. ⁸

Implementing Friedman's proposal to replace Medicare and Medicaid would, Reinhardt said, radically change “distribution of health care in the United States, the distribution of the financial burden of illness, and the socio‐demographic profile of health status” ^{9 (pp. 25‐26)}—surely in ways that would reduce access for those of lesser means. Friedman's conception of efficiency would result in “the rationing of health services and life‐years by price and the individual's ability to pay” when “for all we know, the majority of Americans might well prefer … access to health services by all members of society, on roughly equal terms, regardless of the individual's ability to pay…”. ^{9 (p. 26)}

Then he lays it on: “To suggest that a market‐driven system such as that advocated by Professor Friedman would be more ‘efficient’ than the current American health system with Medicare and Medicaid, or say, the British National Health Services or the Canadian health system, is an idiosyncratic, personal opinion that reflects a personally preferred distributional ethic for health services. It is not the product of ‘economic science’…”. ^{10 (p. xiii)}

Referring to similar writings by economist Richard Epstein, Reinhardt adds, perhaps half humorously, “At the risk of violating the American taboo against class warfare, it is legitimate to observe that virtually everyone who shares Epstein's and Friedman's distributive ethic tends to be rather comfortable ensconced in the upper tiers of the nation's income distribution.”. ^{11 (p. 1447)}

Reinhardt called out Friedman because he was acting not as a scientist, but instead as a political creature who employs seemingly scientific methods to further personally preferred ideological ends. He warned his Princeton students that, “When economists or other policy analysts lapse into their normative mode – when they pretend to be using scientific methods to suggest what ought to be done and what is ‘efficient’ or not – a red warning light should go on in your mind. Chances are that you are being addressed by someone playing politics in the guise of science or by someone insufficiently respectful of the limits of economics as a science.”. ^{10 (p. xiv)}

In further exploring the term “efficiency,” Reinhardt pointed out that its economic meaning is extremely limited and of almost no policy significance. Under certain special conditions—none of which are met in health care—the operation of markets will result in a state of the world called “Pareto optimality,” where it is impossible to make someone better off without making someone else worse off. Such a state of the world is hardly a gold standard, however, for at least three reasons: (a) it has little if any policy significance because the number of public policies that can make someone better off without harming another person is the null set; (b) it completely ignores the ultimate distribution of resources—the main area I focus on here; and (c) it protects the status quo. In one of my favorite non‐Reinhardt quotes of all time, a Nobel Laureate Amartya Sen, stated:

An economy can be [Pareto] optimal … even when some people are rolling in luxury and others are near starvation as long as the starvers cannot be made better off without cutting into the pleasures of the rich. If preventing the burning of Rome would have made Emperor Nero feel worse off, then letting him burn Rome would have been Pareto‐optimal. In short, a society or an economy can be Pareto‐optimal and still be perfectly disgusting. ^{12 (p. 22)}

It is easy to see, then, that people might strongly prefer an “inefficient” allocation of resources that resulted in a better distribution. Reinhardt said that “economists have endowed the term [efficiency] with a precise, technical interpretation that completely divorces it from desirability,” but that “very few lay persons fully understand this fine point.”. ^{7 (p. 306)}

One of Reinhardt's clever expositions of this was his bashing of cost‐benefit analysis (CBA) as practiced by the profession. Everyone knows how CBA works: you examine the total benefits and total costs of alternative projects or policies and pick the one that has the largest difference, or net benefits. Hidden in this method, however, is an assumption that if one person benefits by X dollars, and another is hurt by X‐1 dollars, then society is better off by embarking on the project. But how could that be, particularly in light of the fact that modern economics does not allow us to compare well‐being across different people? The slight‐of‐hand is accomplished by something called the Kaldor‐Hicks criterion, named after two early 20th Century economists. Kaldor, in particular, said that a policy represented a net social gain if losers could be fully compensated for their loss; however, whether they “should be given compensation or not, is a political question on which the economist, qua economist, could hardly pronounce an opinion.”. ^{13 (p. 550)}

Reinhardt called this the “unrequited‐punch‐in‐the‐nose” test—if you get $1000 of jollies from punching me in the nose, and I am willing to pay only $600 for you not to punch me, then society is better off if you punch me—even if I never see a nickel.

I don't think Reinhardt meant to impugn CBA—many projects and policies do indeed help those who are less well‐off—so much as to highlight something much more widespread, and in his view, insidious in health economics: using people's willingness to pay as a measure of social good.

Here's one example of his argument. ⁹ Suppose the Jones family is wealthy and has a healthy baby, and the Smith family is poor with a sickly baby. Then put yourself in Milton Friedman's ideal world, where everyone has a deductible of 30% of their income, meaning, in effect, that all doctor visits are paid out of pocket. Suppose a pediatrician visit cost $40 (he wrote this in 1997), which resulted in 5 visits demanded by the rich Joneses with the healthy baby, but only 3 by the poor Smiths and their sickly baby. To an economist, this is optimal: people's willingness to pay reflects the value they ascribe to the care. It is most important that we understand Reinhardt's thinking about this. He writes:

if one accepts the prevailing after‐tax and after‐transfer distribution of income as the ethically acceptable platform for the market, then it is also ethical to accept as a working proposition that health care given, say, to a child of a corporate executive has a higher social value than similar health care given to the child of a gas station attendant, because the corporate executive probably would be willing to pay more money for that care than would the gas station attendant. ^{14 (p. 976)}

You may be amused that he gave names to these hypothetical people: George and Martha. Reinhardt further writes,

One must wonder how many first‐year students, struggling with their first brush against the complexity of welfare economics, will stop to ponder who George and Martha might be. Would the proposition appeal to them if George were Donald Trump and Martha a near‐poor waitress? ^{14 (p. 982)}

He wrote that in 2001!

Reinhardt further stated that “physicians and public‐health planners may be stunned that anyone would call such an allocation efficient. But reversing the allocation – granting more visits to the sick child and fewer to the healthy child – would be ‘inefficient’, as the term is defined in standard ‘welfare economics’.”. ^{9 (p. 24)}

You probably know the response of an adherent to traditional theory; it would go something like this: If society has a problem, it should re‐allocate incomes through taxes and subsidies so the lower‐income Smiths can afford to bring their baby to the pediatrician as many times as necessary.

This line of reasoning drove Reinhardt crazy for two very good reasons. First, he wasn't buying—not a bit—that the prevailing politics of redistribution of income in society would ever give the Smiths a fighting chance. And second, he wasn't buying that the only good subsidy is a cash subsidy. Economists love cash subsidies because people can spend the money in any way they like, preserving freedom and enshrining their choices in the marketplace as “optimal,” whereby in‐kind subsidies (eg, access to inexpensive medical care) are inefficient because they restrict the individual's choices. The problem is that no society is going to give enough unrestricted cash because donors will worry that their tax contributions likely will not be spent as they like. If one wants to improve the distribution of wealth, the lions share has to be done in‐kind—and of course that's how it's done both in government and charity. But Reinhardt says it better than I, in perhaps my favorite quote:

It can be asked fairly whether so many people could possibly be so wrong on this issue for so many years, while economists have had it right all along, or whether it may possibly be the other way around. ^{14 (p. 977)}

This all may seem a bit in the weeds, but Reinhardt's point was a simple one: we must not consider efficiency and equity separately because no policy is efficient if it ignores equity. He writes:

Indeed, to begin an exploration of alternative proposals for the reform of our health system without first setting forth explicitly, and very clearly, the social values to which the reformed system is to adhere strikes at least this author as patently inefficient: it is a waste of time. Would it not be more efficient merely to explore the relative efficiency of alternative proposals that do conform to widely shared social values? ^{7 (p. 315)}

In sum, Reinhardt wondered how our profession could get its core precepts so wrong. We can see his answer by returning to the issue of physician‐induced demand. Reinhardt writes,

Mastery of the neoclassical framework requires a heavy personal investment on part of the analyst” in a “shared analytic paradigm”—and “among the payoffs to that investment is entrée into a fraternity whose power has derived in good part from the unit of thought forged by this shared analytic paradigm.” Those who question it “are assaulting one of the crucial pillars of the neoclassical framework,” so “it is eminently understandable that the guardians of that framework parry such assaults with vigor. ^{1 (p. 188)}

I don't want you to go away from this talk thinking that theory is all that matters. Theory matters a lot but only as a way of understanding real world phenomena. So how do these lessons about being a good economist, and in particular, appreciating the inseparability of efficiency from equity, apply to current health policy?

As you know, Reinhardt used the world's health care systems as his data‐generating laboratory, providing solid evidence about which countries’ health care systems worked better (in most but not all ways) than ours. At last year's, AcademyHealth Reinhardt lecture, Stuart Altman focused on Germany, Reinhardt's birthplace and probably the national system that he wrote about the most.

I would contend that what nearly all other high‐income countries do differently than ours is to base their health care systems on equity. Not a single one presumes that the prevailing after‐tax income distribution is good enough to ensure that health care will go to everyone that needs it. An excellent example of this is the Netherlands because, more than any other country outside of the US, it relies on managed competition. But there are stark differences between how they do it and how we do it: insurance is universal, nearly all services are covered, patient cost sharing is low, and even low‐income persons have a regular source of care and don't have to wait to obtain specialists’ services. That sort of system makes a lot more sense. Use competition only after you guarantee that those with lesser means are not going to get swept under the rug.

We have been unable to achieve anything remotely like that in the US: Coverage is not universal, key services such as dentistry and long‐term care are severely under‐covered, and cost‐related access barriers are high. Returning to the Netherlands for a moment, the Dutch do face a deductible of about $450 per year, but after that pretty much all costs are covered. ¹⁵ Compare that to our system with its huge deductibles and considerable copayments after the deductible is met. I hope that this lecture has made it clear that making lower‐income persons face these barriers is the opposite of efficient!

Let me end by noting Reinhardt's admonition regarding the training of younger economists, “many of whom no longer seem to explore very carefully the philosophical and ethical underpinnings of their profession and instead concentrate on mere analytic technique.”. ^{7 (p. 313)}

We need to expose not only these younger economists, but the whole health services research community, to Reinhardt's trenchant thoughts and marvelous use of the language, so as to help us all be better economists.